Research ArticleJanus Poly(Vinylidene Fluoride) Membranes with PenetrativePores for Photothermal Desalination

Hao-Hao Yu , Lin-Jiong Yan , Ye-Cheng Shen, Si-Yu Chen , Hao-Nan Li, Jing Yang ,and Zhi-Kang Xu

MOE Key Laboratory of Macromolecular Synthesis and Functionalization, and Key Laboratory of Adsorption and SeparationMaterials & Technologies of Zhejiang Province, Department of Polymer Science and Engineering, Zhejiang University,Hangzhou 310027, China

Correspondence should be addressed to Jing Yang; jing_yang@zju.edu.cn and Zhi-Kang Xu; xuzk@zju.edu.cn

Received 28 October 2019; Accepted 27 January 2020; Published 3 March 2020

Copyright © 2020 Hao-Hao Yu et al. Exclusive Licensee Science and Technology Review Publishing House. Distributed under aCreative Commons Attribution License (CC BY 4.0).

Solar-driven desalination has been considered as a promising technology for producing clean water through an abundant andpollution-free energy source. It is a critical challenge to reasonably design the porous morphology and the thermal managementof photothermal membranes for enabling efficient energy conversion and water production. In this work, a Januspoly(vinylidene fluoride) membrane was fabricated in combination of penetrative pore structure, asymmetric surface wettabilitywith proper thermal management for high-efficiency solar desalination. Highly open and directly penetrative pores achieved bythe two-dimensional solvent freezing strategy are considered to provide direct pathways for water and vapor transportation. Theunique feature of hydrophobic upper layer/hydrophilic lower layer enables the photothermal membranes to self-float on thewater surface and rapidly pump water from the bulk to the surface. The resulting Janus membrane exhibits a satisfactory lightabsorbance as high as 97% and a photothermal conversion efficiency of 62.8% under one-sun irradiation in a direct contactmode. The solar-to-vapor efficiency rises up to 90.2% with the assistance of a thermal insulator adopted beneath. Both the Janusmembrane and the composite setup are able to work efficiently with a high stability in seawater desalination, and theconcentration of ion in condensed water is reduced to below 1 ppm. Therefore, Janus membranes with directly penetrative poresand photothermal surfaces shine a light on the development of high-performance solar evaporators for the practical applicationin solar seawater desalination.

1. Introduction

Freshwater scarcity is one of the most urgent issues for humanbeings in modern society. Solar-driven evaporation for cleanwater production is currently considered as an effective solu-tion to alleviate water shortage because of the ecofriendlyand inexhaustible solar energy [1, 2]. Self-floating evaporatorswith a highly efficient energy conversion and fresh water pro-duction have attracted intensive attention in the solar desali-nation technology due to their strong localization effect onheat conduction at air/water interface [3–10].

Generally, the design philosophy of self-floating solar-driven evaporators with a high efficiency emphasizes an inte-grative optimization of light absorption, thermal manage-ment, and water transportation. First, a high-performancesolar steam generation strongly relies on the photothermal

materials to absorb sunlight as much as possible. Carbon-based materials (e.g., reduced graphene oxide [11, 12], carbonnanotubes [13], carbon black [14, 15]), plasmonic nanopar-ticles [16, 17] (e.g., Al or Au), and super-black polymers(e.g., polypyrrole (PPy) [18, 19], polydopamine (PDA)[20]) have been extensively adopted due to their remarkablesolar-thermal conversion efficiency. Second, an elaboratedthermal management is required, aiming to confine theabsorbed heat on the water surface and then to ensure ther-mal insulation from the bulk water. It can be achieved usingthe evaporator materials with low density [21] and hydro-phobicity [22] or with the aid of porous foams [23, 24]. Thefloating status of the evaporators on the water surface sup-presses the unexpected heat conduction downward to thebulk water. Third, a rapid water and vapor transportation ispreferential to facilitate the water production efficiency,

AAASResearchVolume 2020, Article ID 3241758, 11 pageshttps://doi.org/10.34133/2020/3241758

including water pumping from the bulk body to the evapo-ration surface via the capillary effect and steam escape fromthe photothermal surface to the condenser [25]. To thisend, delivery channels with a low mass transfer resistanceshould be a significant consideration for solar evaporators.Porous substrates containing highly open channels [26]with suitable hydrophilicity [27] play a vital role in theseconcerns above, because they work as not only supportsfor photothermal materials but also channels for waterand vapor transfer. Emerging membranes with directlypenetrative pores across the membrane thickness are pref-erential substrates for the fabrication of self-floating solarevaporators. For instance, a solar evaporator made fromnatural wood achieves a light absorbance of around 99%due to the waveguide effect of columnar pores incorporatedwith plasmonic nanoparticles [28]. Meanwhile, its micro-and nanochannels with a low tortuosity can efficientlypump water from the bottom of the device because of thecapillary wicking effect [28]. Similarly, anodic aluminumoxide (AAO) membranes loaded with Al plasmonic nano-particles on their nanochannel walls feature directly pene-trative pores for water delivery. These pores also functionas efficient light trappers to achieve an impressive lightabsorbance of 99% [17]. More recently, auricle-inspired3D photothermal cones are developed for high-efficiencysolar-driven evaporation. Commercial hydrophilic polyvi-nylidene fluoride (PVDF) membranes are deposited withPPy and then folded into cones to minimize light reflectionand heat loss to bulk water. They exhibit a high solar absor-bance of 99.2% which is higher than that of the plane mem-branes (93%) due to the multiple reflection of sunlightwithin the 3D cones [18]. Nevertheless, the limited process-ability is an intractable issue for natural wood. AAO mem-branes usually suffer from low porosity and mechanicalfragility, and their preparation process may impose a restric-tion on the large-scale fabrication. Moreover, the fabricationof 3D cones involves complicated and multistep processes,including folding and bonding the membrane and sealingthe hole on the apex. Besides, the surface area of 3D cores ismuch larger than the projected area, which could lead to ahigh consumption of photothermal materials. Hence, it ishighly required to develop easy-to-modify and easy-to-process membranes with directly penetrative pores, highporosity, and mechanical robustness for solar-driven photo-thermal evaporators with high performances.

In our previous work, vertically oriented porous mem-branes (VOPMs) were prepared via a thermally inducedphase separation process [29]. The growth of solvent crystalwas controlled from one side to the other side within thePVDF membrane through a well-designed bidirectionalfreezing process. Micron-sized penetrative pores were thusformed across the membrane thickness after removing sol-vent crystals by extraction or sublimation. The resultingVOPMs possess a high porosity (70-80%), and their directlypenetrative pores feature large openings at one side and smallones at the other side, like cones neatly inserted across themembrane. Such distinctive porous morphology makesVOPMs an ideal base substrate for interfacial solar evapora-tors. First, directly penetrative pores are conducive to water

pumping and steam escape. Second, the cone-like pores areable to fully absorb solar energy due to the multiscatteringeffect. Herein, hydrophobic PPy was coated on the inner porewalls of VOPM via chemical vapor deposition polymeriza-tion (CVDP) [30]. Hydrophilic PDA was further depositedon the single side of the PPy-coated VOPM to construct aJanus structure (Figure 1) via the rapid oxypolymerizationof dopamine according to our previous work [31, 32]. Theresulting Janus VOPM floats on the water surface with thehydrophobic layer towards air for absorbing sunlight irradia-tion and the hydrophilic channels facing water for pumpingwater and delivering steam (Figure 1). Such Janus structureresults in a light absorption capacity as high as 97% and anaccelerated water evaporation rate up to 1.08 kg·m-2·h-1under one-sun irradiation in a direct contact mode. The finalsolar-to-vapor conversion efficiency is 62.8%. Furthermore,with the aid of a thermal insulator and a 2D water path[23] settled beneath the Janus VOPM, the evaporation raterises to 1.58 kg·m-2·h-1 and the conversion efficiency reachesup to 90.2%. This composite device is able to operatesmoothly in seawater desalination for several hours, and thecontent of inorganic ions in condensed water is reduced tobelow 1ppm. This study provides a new insight in designinghighly efficient solar evaporators and holds promise for scale-up desalination because of the simple membrane fabricationwith high scalability.

2. Results and Discussion

The fabrication process of VOPMs is well controlled on thebasis of our previously proposed two-dimensional solventfreezing strategy [29]. Dimethyl sulfone (DMSO2) is a goodsolvent for PVDF at a high temperature. It is able to crystal-lize within the polymer matrix in a designed cooling processwith two-dimensional temperature gradients. The nucleationof DMSO2 is caused by the addition of cooling water. Mean-while, the temperature gradient originated from the differ-ence in thermal conductivity between stainless steel andglass promotes the crystal growth along the direction fromthe stainless-steel plate to the glass one. The closer the pre-cursor solution contacts the glass plate, the more crystalliza-tion time is available to form large crystals in the coolingprocess, producing cone-like DMSO2 crystals with thetapered tip towards the stainless-steel plate. After removingDMSO2 crystals by extraction, cone-like and directly pene-trative pores are generated in the thickness direction of thePVDF membrane. The membrane surface contacting thestainless-steel plate shows small pores (diameter ≈ 1 μm)while the other surface touching the glass plate displays largepores (diameter ≈ 3 μm) (Figure 2(a)). The correspondingsurfaces are denoted as S-surface and L-surface, respectively.The as-prepared VOPMs were further used as the skeletonsto fabricate photothermal membranes with a homogeneousPPy coating on inner pore walls via a facile CVDP process[30]. The white nascent VOPM (Figure 2(a)) becomes black(Figure 2(b)) after PPy coating. To further impart the PPy-coated VOPM with hydrophilicity, PDA was decorated onthe S-surface using the single-side-floated deposition trig-gered by CuSO4/H2O2 [31]. The resulting Janus VOPM

2 Research

exhibits a brown side after PDA coating (Figure 2(c)). It isnoteworthy that both postmodification processes have noinfluence on the pore morphology of VOPM, still main-taining an open porous structure without blockage(Figures 2(b) and 2(c)). The chemical structure of eachdecorated layer was analyzed by attenuated total reflectionFourier transform infrared spectroscopy (ATR-FTIR)(Figure S1a) and X-ray photoelectron spectroscopy (XPS)(Figure S1b, Table S1). The decoration of VOPM is alsoreflected by the variation of surface wettability. The watercontact angles (WCA) of nascent VOPM and PPy-coatedone are >120° due to the inherent hydrophobicity of PVDFand PPy (Figure S2a). After the deposition of PDA, theWCA of S-surface decreases to <90° while that of L-surfaceremains >120°, verifying an asymmetric wettability for theJanus VOPM. Moreover, the WCA of S-surface shows acontinuous decline with increasing the deposition time(Figure S2b), suggesting an increasing coverage of PDA onthe surface of pore walls to achieve a hydrophilization onthe single side of VOPM. The hydrophilization depthdetected by laser scanning confocal microscopy (LSCM)[33] is about 15μm at a deposition time of 20min andshows no obvious increasing tendency with prolonging thetime (Figure S3).

The solar absorption ability of photothermal materialsplays a predominate role in the efficiency of the solar vapor

generation process. The solar absorbance (A) of membraneunder one-sun irradiation is calculated by Equation (1):

A = 1 − R − Tð Þ × 100% ð1Þ

where T is the transmittance and R is the diffuse reflectancemeasured from the UV-Vis spectra in the wavelength rangeof 200-800nm (Figure S4). The solar absorbance of nascentVOPM is maintained at a relatively low level (40%-45% inthe visible light range and 45%-70% in the UV region)while that of PPy-coated VOPM reaches >94% in the fullrange of wavelength (Figure 3(a)). It further increases up to~97% after the PDA deposition on the S-surface(Figure 3(a)). Such significantly enhanced solar absorptionis ascribed to the deposition of light absorbers (PPy andPDA) onto the pore walls of VOPM. Besides, it is notableto see that L-surface exhibits a stronger absorption abilitythan S-surface under solar irradiation for the VOPMhomogeneously coated with PPy (Figure S5). Thisphenomenon can be explained by the stronger antireflectivityof large pores compared with small ones under asimulated sunlight with the same intensity (Figure S5).L-surface was, therefore, chosen to be the surface towardssunlight for the solar-driven evaporation in this work. Toassess the light-to-heat ability, the surface temperatures ofphotothermal membranes were probed by a thermal

Pyrrole Dopamine

Figure 1: Schematic illustration to the fabrication of Janus VOPM for photothermal desalination.

3Research

(a)

(b)

(c)

1 cm

1 cm

1 cm

S-surface Cross-section L-surface

S-surface Cross-section L-surface

S-surface Cross-section L-surface

Figure 2: Digital photos and scanning electron microscopy (SEM) images of (a) nascent VOPM, (b) PPy-coated VOPM, and (c) JanusVOPM.

0

20

40

60

80

Tem

pera

ture

(°C)

In dry stateOn water surface

Janus VOPMPPy-coated VOPMNascent VOPM200 400 600 80030

45

60

75

91

94

97

100

Abso

rban

ce (%

)

Wavelength (nm)

Nascent VOPMPPy-coated VOPMJanus VOPM

(a) (b)

Figure 3: Light absorption capacities of various VOPMs. (a) UV-Vis absorption spectra and (b) equilibrium surface temperatures of nascentVOPM, PPy-coated VOPM, and Janus VOPM in a dry state and in a floating state under one-sun irradiation.

4 Research

infrared camera. The surface temperatures of Janus VOPMand PPy-coated VOPM in a dry state are 69:6 ± 3:7°C and65:4 ± 1:2°C under one-sun irradiation, respectively, whichare much higher than that of nascent VOPM (41:9 ± 1:1°C)(Figure 3(b)). In a floating state, both Janus VOPM(41:0 ± 0:9°C) and PPy-coated VOPM (42:5 ± 1:1°C) alsoexhibit higher surface temperatures than the nascentVOPM (29:6 ± 0:3°C) (Figure 3(b)). These results predictthe superiority of Janus VOPMs in interfacial heating forthe solar steam generation.

To investigate the performance of solar vapor generation,different VOPM samples were placed onto the water surfaceunder a simulated solar source which is vertically placedabove the membrane (Figure 4(a)). The weight loss of waterduring the solar evaporation was recorded in real time witha balance below the beaker (Figure S6). Pure waterevaporation under one-sun irradiation and in darknesswere adopted as controls. All VOPM samples are able to

self-float on the water surface benefiting from thehydrophobicity of PVDF and PPy coatings. The evolutionsof water weight loss and corresponding evaporation rateover irradiation time were recorded in Figures 4(b) and4(c), respectively. Under one-sun irradiation, the waterevaporation rate of nascent VOPM slightly increases overtime and reaches 0.48 kg·m-2·h-1 after 65min. This behavioris consistent with the natural water evaporation under thesame condition, suggesting incapability of nascent VOPM toconvert light to heat. In contrast, both Janus VOPM andPPy-coated VOPM significantly accelerate water evaporationwith an ever-increasing rate over time. Their evaporationrates are 1.10kg·m-2·h-1 and 0.91kg·m-2·h-1 after 65min,respectively. The relatively rapid water evaporation usingJanus VOPM is attributed to the hydrophilized watertransportation pathway coated with PDA which speeds upthe water delivery during the solar evaporation. It is worthyto note that the hydrophilization depth of Janus VOPM

0 20 40 60

0.0

0.4

0.8

1.2

1.6

Wei

ght l

oss (

kg. m

-2)

Time (min)

With thermal insulator under 1 sun Without thermal insulator under 1 sunWith thermal insulator in darknessWithout thermal insulator in darkness

0 10 20 30 40 50 60

0.0

0.4

0.8

1.2

Wei

ght l

oss (

kg. m

-2)

Time (min)

Janus VOPM under 1 sunPPy-coated VOPM under 1 sunNascent VOPM under 1 sunPure water under 1 sunJanus VOPM in darknessPure water in darkness

(b)

(e)

0.0

0.4

0.8

1.2

1.6

0 20 40 600.0

0.4

0.8

1.2

1.6

Evap

orat

ion

rate

(kg.

m-2

. h-1

)Ev

apor

atio

n ra

te (k

g.m

-2. h

-1)

(a)

Sunlight

Vapor

Water

(d)

Thermal insulator

2D waterpath

0 20 40 60Time (min)

With thermal insulator under 1 sun Without thermal insulator under 1 sunWith thermal insulator in darknessWithout thermal insulator in darkness

(f)

Time (min)

Janus VOPM under 1 sunPPy-coated VOPM under 1 sunNascent VOPM under 1 sunPure water under 1 sunJanus VOPM in darknessPure water in darkness

(c)

Figure 4: Setups and their performances of solar vapor generation. (a) Direct contact mode for solar evaporation. (b) Weight loss and (c)evaporation rate of water as a function of irradiation time using different VOPMs in a direct contact mode. (d) Composite deviceequipped with a thermal insulator and an absorbent paper for solar evaporation. (e) Weight loss and (f) evaporation rate of water as afunction of irradiation time using Janus VOPMs with or without the thermal insulator.

5Research

remains constant with prolonging the deposition time(Figure S3) as mentioned above. This result provides areasonable explanation for the stable water evaporation ratesachieved by Janus VOPMs with different deposition time(Figure S7). As a control, the evaporation rate of JanusVOPM in darkness is almost invariable over time until itreaches 0.10kg·m-2·h-1 after 65min, which is as same as thepure water evaporation (Figures 4(b) and 4(c)). The solar-to-vapor conversion efficiency (η) was further calculated byEquation (2) [3]:

η = _m − _m0ð ÞHv

CoptP0ð2Þ

where _m is the evaporation rate, _m0 is the natural evaporationrate without irradiation, Hv is the enthalpy of liquid-vaporphase change (~2260kJ·kg-1), Copt is the optical concentration,and P0 is the power density of one-sun irradiation (1kW·m-2).

The solar-to-vapor conversion efficiency of natural waterevaporation under one sun is only 23.8%, while those of PPy-coated VOPM and Janus VOPM reach up to 54.6% and62.8%, respectively. Although this efficiency assisted withJanus VOPM is 2.6 times higher than that of natural waterevaporation, it is still much lower than the light absorbanceof the membrane (>97%). The reason for this loss ofabsorbed solar energy is ascribed to the intrinsic limitation ofthe direct contact mode (Figure 4(a)) in which a significantportion of absorbed energy unavoidably dissipates throughthe bulk water. To suppress the heat loss, a composite deviceequipped with a thermal insulator layer and a 2D water pathwas adopted [23], as illustrated in Figure 4(d). A hydrophobicpolymer foam (closed-cell polyurethane foam, thickness ≈ 20mm, thermal conductivity ≈ 0:034W·m-1·K-1) was wrappedby an absorbent paper, and the Janus VOPM was placed onthe top. The downward heat conduction can be dramaticallyreduced by the thermally insulated foam. Meanwhile, wateris continuously pumped to the light absorber layer throughthe absorbent paper. In this way, both efficient water supplyand depressed heat loss can be achieved simultaneously. Withthis device, the equilibrium surface temperature under one-sun irradiation reaches 47:0 ± 0:44°C, much higher than thatusing the direct contact mode (41:0 ± 0:9°C). Therefore, theevaporation rate strikingly increases up to 1.58 kg·m-2·h-1 after35min and remains relatively stable over time (Figures 4(e)and 4(f)), resulting in an enhancement of solar-to-vapor con-versionefficiencyup to90:2 ± 3:1%underone-sun irradiation.

Table 1 summarizes the performances of solar evapora-tors using recently reported photothermal membranes. TheJanus VOPM with penetrative pores in this work achieves ahigh evaporation rate and solar-to-vapor conversion effi-ciency compared with others, especially when it is equippedwith a thermal insulator. It is also worth stressing that theJanus VOPM shows a higher solar-to-vapor conversion effi-ciency than membranes with a bicontinuous structure viaphase inversion [18] and fibrous membranes via electrostaticspinning [15]. It is the microsized cone-like pores which aredirectly penetrative through the membrane thickness thatpromote the light harvesting and improve the vapor escape

rate due to the capillary wicking effect, generating watersteam more efficiently.

In view of the requirements for effective and efficient solardesalination via solar-driven steam generation, the seawaterevaporation performances using Janus VOPM were furtherinvestigated. With the aid of a thermal insulator, the seawaterevaporation remains stable with a rate of 1.5~1.6 kg·m-2·h-1throughout a continuous operation for 7h, indicating a highsolar-to-vapor conversion efficiency and a long-term durabil-ity for seawater desalination (Figure 5(a)). The stability of thesystem was further investigated by a cyclic testing in which theoperation time for each cycle is 1h. The evaporation rates arealmost invariable during 10 cycles, maintaining a high level of>1.5 kg·m-2·h-1 (Figure S8). Furthermore, a concentrated NaClsolution (20%) was used to test the performance of solardesalination (Figure S9a). The evaporation rate increases atfirst since the surface temperature is getting higher due tothe light-to-heat conversion. However, it shows a decline inhalf an hour because of the salt crystal deposition on themembrane surface. The asymmetric wettability of the JanusVOPM can effectively prevent the salting-out effect since thesalts accumulated onto the hydrophilic PDA layer can bequickly dissolved through the continuous water pumping atthe initial stage. However, the salt concentration increasesgreatly in the 2D water path (absorbent paper), andeventually, the salt crystals are formed at the interfacebetween the Janus VOPM and the absorbent paper in a fewhours (Figure S9b insert). These salt crystals would intensifythe light reflection and block the water or vapor channels,resulting in a depression of evaporation rate (Figure S9b).What is more, the accumulated salts would bring amechanical stress to the device and even damage the JanusVOPM. On the contrary, the Janus VOPM in a directcontact mode without a thermal insulator is able to operatesmoothly for as long as 12h without an obvious decrease inevaporation rate (Figure S10), indicating that the membranewith directly penetrative pores and the Janus structureefficiently prevent the salt blocking. Consequently, wesuggest that the Janus VOPM with a direct contact mode onwater surface is more suitable to achieve a long-term stabilityfor the solar desalination of highly brackish water. However,for desalination of seawater or of other systems withrelatively low salt concentration, the Janus membrane witha thermal insulator is preferred to greatly improve theefficiency of water evaporation.

Eventually, the ion concentrations in seawater andcondensed water were measured by inductively coupledplasma emission spectroscopy (ICP) to evaluate the qualityof desalinated fresh water through the solar-driven evapo-ration. The concentrations of all inorganic ions decreasesharply below 1ppm after the evaporation-condensationprocess (Figure 5(b)). The rejections to these ions are ashigh as 99.8%, which is fully compliant with the potablewater standard issued by the World Health Organization.

3. Conclusion

In summary, we have developed a facile strategy to fabricateJanus photothermal membranes with directly penetrative

6 Research

Table1:Solarvapo

rgeneration

performancesof

differentmem

branes

underon

e-sunirradiation(C

opt=

1)repo

rted

intheliteratures.

Materials

Morph

ology

Mod

eLightabsorbance

Evapo

ration

rate

Solar-vapo

rconversion

efficiency

Ref.

(%)

(kg·m

-2·h-1)

(%)

AAO

loaded

withAln

anop

articles

Vertically

alignedpo

res

Directcontact

96.0

0.93

57.7

[17]

CuS-coatedPEmem

brane

Intercon

nected

macropo

res

Directcontact

93.0

1.02

63.9

[21]

Hierarchicalcop

per-silicon

nano

wirepo

rous

mem

brane

Intercon

nected

macropo

res

Directcontact

93.8

0.81

50.9

[34]

Withthermalinsulator

93.8

1.37

86.0

PPy-coated

hydrop

hilic

PVDFmem

brane

Intercon

nected

macropo

res

Directcontact

93.0

0.92

54.3

[18]

Folded

into

cones

99.2

1.70

93.8

Electrospinning

CB/PMMA-PAN

Janu

sabsorber

Intercon

nected

macropo

res

Directcontact

97.0

0.92

51.0

[15]

Withthermalinsulator

97.0

1.30

72.0

Janu

sVOPM

Vertically

alignedpo

res

Directcontact

97.0

1.08

62.8

Thiswork

Withthermalinsulator

97.0

1.58

90.2

7Research

cone-like pores for solar steam generation. Hydrophobicsuper-black PPy and hydrophilic PDA are successivelycoated onto the opposing surfaces of PVDF VOPM to con-struct both photothermal function and asymmetric wettabil-ity. The obtained Janus photothermal membranes show ahigh light absorption of over 97% and are able to convertsunlight to heat effectively. The water evaporation rate ofthe Janus VOPM can be accelerated up to 1.10 kg·m-2·h-1under one-sun irradiation using a direct contact mode, andthe solar-vapor conversion efficiency is calculated to be62.8%. With the assistance of a thermal insulator, the waterevaporation rate of Janus VOPM increases to 1.58 kg·m-2·h-1 and the solar-vapor conversion efficiency rises to 90.2%.The setup with a thermal insulator is able to operate contin-uously in seawater, and the contents of salt ions in the con-densed water decrease to below 1ppm. The present strategyprovides a way for rational design and controllable construc-tion of high-performance photothermal membranes, show-ing great potential applications in seawater desalination,waste water treatment, and solution concentration.

4. Materials and Methods

4.1. Experimental Design. Thermally induced phase separa-tion is used to prepare PVDF membranes with directly pen-etrative pores via a two-dimensional solvent freezingtechnique. Subsequently, PPy is coated on one surface ofthe membrane via chemical vapor deposition polymeriza-tion. The other side of the membrane is further hydrophi-lized with PDA via mussel-inspired deposition. These twoprocedures above are aimed at constructing both photother-mal function and asymmetric wettability for Janus photo-thermal membranes. UV-Vis spectroscopy and in situtemperature measurement under a simulated solar irradia-tion are used to investigate photothermal conversion capabil-ity of the Janus photothermal membranes. The total

efficiency of solar-vapor generation can be assessed byrecording the water weight loss in real time.

4.2. Reagents and Materials. Commercially available prod-ucts of PVDF with different molecular weights were obtainedfrom Solvay Solexis (Belgium) (Mn = 110,000 g/mol, Solef6010) and from Shanghai 3F NewMaterials Co., Ltd. (China)(Mn = 500,000 g/mol, FR904) and were dried under a vac-uum at 60°C for 6 h before use. DMSO2 (99.97%) was pur-chased from Dakang Chemicals Co., Ltd. (China). Pyrrole(GC grade) was purchased from Aladdin Chemical Co.,Ltd. (China). Ammonium persulfate (APS) (AR grade),CuSO4·5H2O (AR grade), and H2O2 (30%) were bought fromSinopharm Chemical Reagent Co., Ltd. (China). Dopaminehydrochloride (98%) was purchased from Sigma-Aldrich(China). Sodium fluorescein (AR grade) was purchased fromMacklin Biochemical Co., Ltd. (China). Closed-cell poly-urethane foam was bought from Hebei Duken Energy Sav-ing Technology Co., Ltd. (China).

4.3. Fabrication of VOPM. The fabrication of VOPM is basedon our previously reported method [29]. DMSO2 was used asa crystallizable solvent to prepare a homogenous solution ofPVDF. Commercially available PVDF powders of Solef6010 and FR904 with different molecular weights were mixedat a ratio of 7/3 (w/w) and dissolved in DMSO2 at 180°C witha polymer concentration of 22.5wt%. After degassing toremove air bubbles, the solution was then sealed in a pre-heated (160°C) home-made mold consisting of a stainless-steel plate (thickness = 1mm) and a glass plate(thickness = 3mm). A Teflon film with a square opening(10 × 10 cm) and a thickness of 200μm was inserted betweentwo plates to reserve the polymer solution. The plates werethen clamped together by clips tightly. Subsequently, themold was put in an oven at 160°C for several minutes tomaintain a constant temperature. After being taken out from

0 60 120 180 240 300 360 420

0

3

6

9

12

15

0.0

0.5

1.0

1.5

2.0

Time (min)

Weight lossEvaporation rate

0

20

40

60

80

100105

104

103

102

101

100

10-1

10-2

B3+Ca2+Mg2+K+

Con

cent

ratio

n (p

pm)

Before desalinationAfter desalination

Na+

Reje

ctio

n (%

)

(a) (b)

Wei

ght l

oss (

kg. m

-2)

Evap

orat

ion

rate

(kg.

m-2

. h-1

)

Figure 5: Seawater desalination performances of a Janus VOPM with a thermal insulator. (a) Weight loss and evaporation rate of water as afunction of irradiation time under one-sun illumination. (b) Ion concentrations of seawater before and after desalination. Red balls refer to thecalculated rejection ratios of ions.

8 Research

the oven, the mold was vertically placed in a reservoir withthe addition of water (30°C) at a rate of 1.7mm/s. After thefull concretion, the newborn membrane was taken out fromthe mold and immersed into deionized water to extractDMSO2. The membrane was then washed with ethanoland hexane in sequence and dried under a vacuum at 60°Cfor 24 h.

4.4. Preparation of PPy Layer. The as-prepared hydrophobicVOPM was cut into circular pieces and immersed succes-sively in ethanol and in 0.5M of APS solution for 10min.After removing the surface liquid by filter papers, the wetmembrane was placed (L-surface up) in a sealed container(2 L) which was preheated to 50°C. A beaker filled with20μL of pyrrole was then located beside in the container toinitiate the CVDP reaction. The PPy-deposited membranewas taken out from the container in 30min, washed withdeionized water, and dried under a vacuum at 60°C for 6 h.

4.5. Single-Side Deposition with PDA Coating. The single-sidedecoration of PPy-coated VOPM was conducted by themussel-inspired deposition accelerated by CuSO4/H2O2reported in our previous work [31]. Dopamine hydrochlo-ride (2mg/mL) and CuSO4·5H2O (1.25mg/mL) were dis-solved in a Tris buffer solution (pH = 8:5, 50mM). ThePPy-coated VOPM was prewetted with ethanol and thenfloated on the solution surface. An aqueous solution ofH2O2 (30%) was added into the solution to immediately trig-ger the oxypolymerization of dopamine. The single-sidemodified membrane was taken out in a few minutes andwashed with deionized water overnight. The resulting Janusmembrane was then dried under a vacuum for over 6 h. Tomeasure the modification depth, the Janus membranes wereprewetted by 80% ethanol solution and then immersed inan aqueous solution of sodium fluorescein (1.0mg·mL-1) tostain the hydrophilic PDA layer. The membranes were takenout in 5min and washed with deionized water for the LSCMimaging [33].

4.6. Characterization. Field Emission Scanning ElectronMicroscopy (FE-SEM) images were recorded with HitachiS-4800 (Japan). UV-Vis spectroscopy in a reflection andtransmission model was conducted by a UV-2450 spectrom-eter equipped with an integration sphere (SHIMADZU,Japan). The surface wettability was assessed by the measure-ment of WCA on a Meter A-200 system (MAIST VisionInspection & Measurement Co. Ltd., China). The surfacetemperatures of samples were recorded and analyzed by aniPhone equipped with a thermal infrared camera accessory(FLIR ONE PRO, Flir System. Inc., USA). The concentra-tions of ions in seawater and condensed water were measuredby ICP (Varian 730-ES). The seawater was diluted 5000 timesto meet the measuring range of the ICP testing.

4.7. Water Evaporation Performance Measurement. The sim-ulated solar irradiation was realized with a Xenon lamp (PLXQ500W, Changzhou Hongming Instrument TechnologyCo., Ltd., China). A glass baker was filled with water to aheight of about 7 cm. A circular sample of membrane witha diameter of 4 cm was then placed on the water surface.

The simulated sunlight from the lamp perpendicularly irradi-ated, and the distance between the membrane and the Xenonlamp was strictly controlled to obtain an intensity of solarirradiation of 1 kW · m−2 with the help of a solar powermeter. The foam was also cut into a circular sample with adiameter of 4 cm and was wrapped with an absorbent paper.The water content under the composite device was reducedto make sure the intensity of solar irradiation to the mem-brane surface equaled to 1 kW · m−2. The mass change ofthe system was recorded every 10min by an electronic bal-ance. All the data were recorded at ambient temperature of18 ± 2°C and relative humidity of 60 ± 10%. The evaporationrate ( _m) was calculated by Equation (3):

_m = ΔmS × Δt

ð3Þ

where Δm is the mass change of the system, S is the surfacearea of the membrane, and Δt is the irradiation time. Thedurability of the membrane was evaluated by 10 cyclic testsof evaporation. Each cycle included one-sun irradiation for1 h and refreshing with deionized water for 0.5 h.

Data Availability

All data needed to evaluate the conclusions in the paper arepresent in the paper and the Supplementary Materials. Addi-tional data related to this paper may be requested from theauthors.

Conflicts of Interest

The authors declare that there is no conflict of interestregarding the publication of this article.

Authors’ Contributions

Z.-K. Xu and J. Yang supervised the research; H.-H. Yu, L.-J. Yan, Y.-C. Shen, and S.-Y. Chen conducted the experi-ments; H.-H. Yu, J. Yang, and Z.-K. Xu analyzed the data;J. Yang and H.-H. Yu wrote the manuscript; L.-J. Yan,H.-H. Yu, and H.-N. Li completed the experiments sug-gested from reviewers. Hao-Hao Yu and Lin-Jiong Yancontributed equally to this work.

Acknowledgments

This work was supported by the National Natural ScienceFoundation of China (Grant no. 51673166 and no.51803180).

Supplementary Materials

Figure S1: (a) ATR-FTIR and (b) XPS spectra of nascentVOPM and Janus VOPM with L- and S-surface for thedetection. Figure S2: (a) WCA on the L- and S-surfaceof nascent VOPM, PPy-coated VOPM, and Janus VOPM(deposition time for PDA coating is 20min); (b) WCAon the L- and S-surface of Janus VOPM as a function ofdeposition time. Figure S3: 3D LSCM images of Janus

9Research

VOPMs constructed with various deposition times of PDA(from 0min to 60min). Figure S4: (a) diffuse reflectancespectra and (b) transmission spectra of nascent VOPMand differently treated VOPMs. Figure S5: UV-Vis absorp-tion spectra of PPy-coated VOPM with S- and L-surfacetowards light, respectively. The inserts are schematic illus-trations to the multiscattering effect of cone-like pores.Figure S6: schematic illustration to the setup for themeasurement of evaporation rate in a real time under asimulated solar source. Figure S7: water evaporation ratesof Janus VOPMs with different deposition time of PDAunder one-sun illumination. Figure S8: evaporation ratevariations of a Janus VOPM with a thermal insulatorunder 10 cycles of solar desalination for seawater (the con-dition of one-sun irradiation for 1 h is used for eachcycle). Figure S9: (a) weight loss and (b) evaporation rateof water as a function of time using a Janus VOPM witha thermal insulator for the desalination of seawater andNaCl solution (20wt%) under one-sun illumination. FigureS10: weight loss and evaporation rate of water as a func-tion of time using a Janus VOPM without a thermal insu-lator for the desalination of NaCl solution (20wt%) underone-sun illumination. Table S1: surface element composi-tions (calculated from XPS results) of nascent VOPMand Janus VOPM with L- and S-surface for the detection.(Supplementary Materials)

References

[1] S. Chu and A. Majumdar, “Opportunities and challenges for asustainable energy future,” Nature, vol. 488, no. 7411, pp. 294–303, 2012.

[2] N. S. Lewis, “Research opportunities to advance solar energyutilization,” Science, vol. 351, no. 6271, article aad1920, 2016.

[3] P. Tao, G. Ni, C. Song et al., “Solar-driven interfacial evapora-tion,” Nature Energy, vol. 3, no. 12, pp. 1031–1041, 2018.

[4] H. Sharon and K. S. Reddy, “A review of solar energy drivendesalination technologies,” Renewable and Sustainable EnergyReviews, vol. 41, pp. 1080–1118, 2015.

[5] M. Chandrashekara and A. Yadav, “Water desalination systemusing solar heat: a review,” Renewable and Sustainable EnergyReviews, vol. 67, pp. 1308–1330, 2017.

[6] A. Politano, P. Argurio, G. Di Profio et al., “Photothermalmembrane distillation for seawater desalination,” AdvancedMaterials, vol. 29, no. 2, article 1603504, 2017.

[7] J. Wu, K. R. Zodrow, P. B. Szemraj, and Q. Li, “Photothermalnanocomposite membranes for direct solar membrane distilla-tion,” Journal of Materials Chemistry A, vol. 5, no. 45,pp. 23712–23719, 2017.

[8] Y. Li, X. Cui, M. Zhao et al., “Facile preparation of a robustporous photothermal membrane with antibacterial activityfor efficient solar-driven interfacial water evaporation,” Jour-nal of Materials Chemistry A, vol. 7, no. 2, pp. 704–710,2019.

[9] H. Ghasemi, G. Ni, A. M. Marconnet et al., “Solar steam gen-eration by heat localization,” Nature Communications, vol. 5,no. 1, article 4449, 2014.

[10] Z. Wang, Y. Liu, P. Tao et al., “Bio-inspired evaporationthrough plasmonic film of nanoparticles at the air-water inter-face,” Small, vol. 10, no. 16, pp. 3234–3239, 2014.

[11] P. Zhang, J. Li, L. Lv, Y. Zhao, and L. Qu, “Vertically alignedgraphene sheets membrane for highly efficient solar thermalgeneration of clean water,” ACS Nano, vol. 11, no. 5,pp. 5087–5093, 2017.

[12] T. Chen, S. Wang, Z.Wu et al., “A cake making strategy to pre-pare reduced graphene oxide wrapped plant fiber sponges forhigh-efficiency solar steam generation,” Journal of MaterialsChemistry A, vol. 6, no. 30, pp. 14571–14576, 2018.

[13] Z. Yin, H. Wang, M. Jian et al., “Extremely black verticallyaligned carbon nanotube arrays for solar steam generation,”ACS Applied Materials & Interfaces, vol. 9, no. 34,pp. 28596–28603, 2017.

[14] Y. Jin, J. Chang, Y. Shi, L. Shi, S. Hong, and P. Wang, “Ahighly flexible and washable nonwoven photothermal clothfor efficient and practical solar steam generation,” Journalof Materials Chemistry A, vol. 6, no. 17, pp. 7942–7949,2018.

[15] W. Xu, X. Hu, S. Zhuang et al., “Flexible and salt resistantJanus absorbers by electrospinning for stable and efficient solardesalination,” Advanced Energy Materials, vol. 8, no. 14, article1702884, 2018.

[16] K. Bae, G. Kang, S. K. Cho, W. Park, K. Kim, and W. J. Padilla,“Flexible thin-film black gold membranes with ultrabroad-band plasmonic nanofocusing for efficient solar vapour gener-ation,” Nature Communications, vol. 6, no. 1, 2015.

[17] L. Zhou, Y. Tan, J. Wang et al., “3D self-assembly of alumin-ium nanoparticles for plasmon-enhanced solar desalination,”Nature Photonics, vol. 10, no. 6, pp. 393–398, 2016.

[18] Y. Wang, C. Wang, X. Song et al., “Improved light-harvestingand thermal management for efficient solar-driven water evap-oration using 3D photothermal cones,” Journal of MaterialsChemistry A, vol. 6, no. 21, pp. 9874–9881, 2018.

[19] B. Yu, J. Duan, J. Li et al., “All-day thermogalvanic cells forenvironmental thermal energy harvesting,” Research,vol. 2019, article 2460953, pp. 1–10, 2019.

[20] C. Zhang, M.-B. Wu, B.-H. Wu, J. Yang, and Z. K. Xu, “Solar-driven self-heating sponges for highly efficient crude oil spillremediation,” Journal of Materials Chemistry A, vol. 6,no. 19, pp. 8880–8885, 2018.

[21] M. Shang, N. Li, S. Zhang et al., “Full-spectrum solar-to-heatconversion membrane with interfacial plasmonic heating abil-ity for high-efficiency desalination of seawater,” ACS AppliedEnergy Materials, vol. 1, no. 1, pp. 56–61, 2018.

[22] L. Zhang, B. Tang, J. Wu, R. Li, and P. Wang, “Hydrophobiclight-to-heat conversion membranes with self-healing abilityfor interfacial solar heating,” Advanced Materials, vol. 27,no. 33, pp. 4889–4894, 2015.

[23] X. Li, W. Xu, M. Tang et al., “Graphene oxide-based efficientand scalable solar desalination under one sun with a confined2D water path,” Proceedings of the National Academy of Sci-ences of the United States of America, vol. 113, no. 49,pp. 13953–13958, 2016.

[24] Q. Zhang, X. Xiao, G. Wang et al., “Silk-based systems forhighly efficient photothermal conversion under one sun: por-tability, flexibility, and durability,” Journal of Materials Chem-istry A, vol. 6, no. 35, pp. 17212–17219, 2018.

[25] M. Zhu, Y. Li, G. Chen et al., “Tree-inspired design for high-efficiency water extraction,” Advanced Materials, vol. 29,no. 44, article 1704107, 2017.

[26] Y. Ito, Y. Tanabe, J. Han, T. Fujita, K. Tanigaki, and M. Chen,“Multifunctional porous graphene for high-efficiency steam

10 Research

generation by heat localization,” Advanced Materials, vol. 27,no. 29, pp. 4302–4307, 2015.

[27] S. Yu, Y. Zhang, H. Duan et al., “The impact of surface chem-istry on the performance of localized solar-driven evaporationsystem,” Scientific Reports, vol. 5, no. 1, article BFsrep13600,2015.

[28] M. Zhu, Y. Li, F. Chen et al., “Plasmonic wood for High-Efficiency solar steam generation,” Advanced Energy Mate-rials, vol. 8, no. 4, article 1701028, 2018.

[29] H.-Q. Liang, K.-J. Ji, L.-Y. Zha, W. B. Hu, Y. Ou, and Z. K. Xu,“Polymer membranes with vertically oriented pores con-structed by 2D freezing at ambient temperature,” ACS AppliedMaterials & Interfaces, vol. 8, no. 22, pp. 14174–14181, 2016.

[30] L. M. Santino, S. Acharya, and J. M. D'Arcy, “Low-temperaturevapour phase polymerized polypyrrole nanobrushes for super-capacitors,” Journal of Materials Chemistry A, vol. 5, no. 23,pp. 11772–11780, 2017.

[31] C. Zhang, Y. Ou, W.-X. Lei, L. S. Wan, J. Ji, and Z. K. Xu,“CuSO4/H2O2-induced rapid deposition of polydopaminecoatings with high uniformity and enhanced stability,” Ange-wandte Chemie, vol. 128, no. 9, pp. 3106–3109, 2016.

[32] H.-C. Yang, M.-B. Wu, J. Hou, S. B. Darling, and Z.-K. Xu,“Nanofilms directly formed on macro-porous substrates formolecular and ionic sieving,” Journal of Materials ChemistryA, vol. 6, no. 7, pp. 2908–2913, 2018.

[33] J. Yang, H.-N. Li, Z.-X. Chen, A. He, Q.-Z. Zhong, andZ.-K. Xu, “Janus membranes with controllable asymmetricconfigurations for highly efficient separation of oil-in-wateremulsions,” Journal of Materials Chemistry A, vol. 7, no. 13,pp. 7907–7917, 2019.

[34] X. Song, H. Song, N. Xu et al., “Omnidirectional and effectivesalt-rejecting absorber with rationally designed nanoarchitec-ture for efficient and durable solar vapour generation,” Journalof Materials Chemistry A, vol. 6, no. 45, pp. 22976–22986,2018.

11Research